HRV Demystified: What the Number Actually Means

What HRV actually measures.

Your heart does not beat like a metronome. The time between consecutive beats — the R-R interval, named after the R peaks on an ECG trace — varies from beat to beat. HRV quantifies that variation. A heart beating at 60 bpm is not firing once per second with mechanical precision. It might fire at 0.95 seconds, then 1.04, then 0.98. That irregularity is the signal. And counterintuitively, more irregularity is generally better.

The variation comes from two competing branches of your autonomic nervous system. The sympathetic branch — fight or flight — speeds the heart and reduces variability. The parasympathetic branch — rest and digest, mediated by the vagus nerve — slows the heart and increases variability. HRV is a window into the tug-of-war between these two systems. When parasympathetic influence dominates, beat-to-beat intervals vary more. When sympathetic tone is elevated — from stress, under-recovery, illness, or overtraining — the heart locks into a tighter, less variable rhythm.

Higher HRV generally means greater parasympathetic influence, which correlates with better recovery capacity. But “higher is better” has limits. Parasympathetic rebound after exhaustive training can temporarily inflate HRV even when the body is still recovering.^[2] A single high reading the morning after a brutal session does not mean you are recovered. It may mean your autonomic nervous system is overcompensating. HRV is a proxy for autonomic balance — not a direct readout of fitness, readiness, or health.

RMSSD and why it won.

There are dozens of ways to quantify heart rate variability. The consumer wearable industry settled on one: RMSSD — the root mean square of successive differences. It calculates the square root of the average squared differences between consecutive R-R intervals. What matters is what it reflects: short-term, beat-to-beat variation driven primarily by parasympathetic activity.^[1]

HRV metrics fall into two broad categories. Time-domain metrics like RMSSD and SDNN (standard deviation of normal-to-normal intervals) work directly with the intervals themselves. Frequency-domain metrics like the LF/HF ratio decompose the signal into spectral bands — low frequency and high frequency — and attempt to separate sympathetic from parasympathetic contributions. For a decade, the LF/HF ratio was popular in research. Then the field largely moved on. Buchheit argued persuasively that the LF band reflects too many overlapping physiological processes to be a clean marker of sympathetic activity, making the ratio unreliable for practical monitoring.^[3]

RMSSD won for a practical reason: it works with short recordings. SDNN requires longer measurement windows — typically five minutes or more — to stabilize. RMSSD produces reliable values from recordings as short as 60 seconds.^[1] That matters enormously for consumer devices. Nobody is lying still for ten minutes every morning. One to two minutes is the realistic window, and RMSSD fills it well.

Hardware: chest strap vs. wrist.

An ECG chest strap measures the electrical activity of the heart directly. It detects the actual R-R intervals with millisecond precision. This is the gold standard for HRV measurement, and it is what most research protocols use. If accuracy per reading is your priority, a Polar H10 or similar chest strap with an HRV app is the cleanest signal you will get outside a clinical setting.

Wrist-based wearables — Apple Watch, Garmin, Whoop, Oura — use a different technology: photoplethysmography (PPG). A green LED shines into your skin and measures changes in blood volume at the capillary level. From that pulse signal, the device infers the timing between heartbeats. This is pulse rate variability, an approximation of HRV. It is not the same measurement. PPG is affected by wrist fit, skin tone, motion artifacts, and the distance between the sensor and the vasculature. During movement, accuracy degrades substantially.

Apple Watch overnight recordings partially solve the motion problem. When you are asleep and still, PPG artifacts drop and the signal quality improves. Overnight averages from wrist devices are more reliable than waking spot checks. But the single most important principle is this: consistency of device and timing matters more than absolute accuracy. A chest strap at 6 AM every morning and an Apple Watch overnight average are both valid approaches. A chest strap on Monday, a wrist check on Wednesday, and an overnight average on Friday are not. Pick one method. Use it the same way every time. The trend is the signal, not any individual number.

The morning measurement protocol.

If you are taking a manual HRV reading (not relying on overnight wearable data), the protocol is the measurement. Change the conditions and you change the number. The standard morning protocol exists for a reason: it controls for the variables that would otherwise drown the signal in noise.

• •Within five minutes of waking. Before you stand up, if possible.
• •Before coffee. Caffeine is a sympathetic activator. It will lower your reading.
• •Before checking your phone. The email from your boss, the news headline, the Instagram scroll — each triggers a cortisol response that shifts autonomic balance toward sympathetic dominance.
• •Supine or seated — pick one and never switch. Standing introduces orthostatic stress that drops HRV. Supine and seated produce different baselines. Both are valid. Mixing them is not.
• •One to two minutes minimum recording length. Shorter recordings increase measurement noise.

Each rule traces back to a specific confounder. Standing up engages the baroreceptor reflex and shifts blood pooling, immediately altering heart rate dynamics. Caffeine binds adenosine receptors and triggers catecholamine release. Psychological stress from phone use activates the hypothalamic-pituitary-adrenal axis within seconds. Body position changes the gravitational load on the cardiovascular system. These are not minor effects — any of them can swing RMSSD by 10–20 ms or more, which in many people represents the entire range between a “recovered” and “suppressed” reading.

What medications do to HRV.

Any medication that affects heart rate or autonomic function will shift your HRV baseline. This does not make HRV tracking useless on medication — it means your baseline is medication-adjusted, and you should compare only to your own trend, not to population norms or someone else’s numbers.

Beta-blockers

Beta-blockers reduce sympathetic influence on the heart, which increases parasympathetic tone and raises HRV. On paper, this looks like improved recovery. In practice, it can mask overtraining. If your HRV stays elevated because the medication is holding sympathetic activation in check, you lose the early warning signal that you are digging a recovery hole. Anyone using HRV-guided training decisions while on a beta-blocker should weight subjective recovery markers — sleep quality, mood, motivation, joint soreness — more heavily than the number itself.

Stimulants and caffeine

Stimulants acutely increase sympathetic nervous system activity, which suppresses HRV. This is why the morning protocol demands measurement before coffee. Even 100 mg of caffeine — a small cup of coffee — can drop RMSSD measurably within 20–30 minutes. Prescription stimulants (amphetamines, methylphenidate) have a more pronounced and longer-lasting effect.

TRT

Direct evidence linking testosterone replacement therapy to HRV changes is limited. Testosterone does not act on the heart’s conduction system the way beta-blockers or stimulants do. However, adequate testosterone levels may improve sleep architecture — particularly deep sleep duration — and improved sleep quality is associated with higher morning HRV. This is physiological inference, not direct trial data. If your HRV improves after starting TRT, the mechanism is likely indirect: better sleep, improved mood, reduced systemic inflammation, and more consistent training.

GLP-1 agonists

Weight loss from GLP-1 medications may improve HRV over time. Excess adiposity is associated with chronic low-grade inflammation and elevated sympathetic tone, both of which suppress HRV. As body fat decreases and metabolic health markers improve, autonomic balance may shift toward greater parasympathetic influence. Acutely, however, GI distress — nausea, vomiting, disrupted sleep from discomfort — can temporarily suppress HRV. This is physiological reasoning based on known mechanisms, not direct RCT evidence linking semaglutide to HRV outcomes.

Age, fitness, and your baseline.

HRV declines with age. This is well-documented, consistent across populations, and not pathological.^[1] A 50-year-old with a morning RMSSD of 30 ms is not “unhealthy” compared to a 25-year-old reading 65 ms. They are in different decades of life with different autonomic baselines. The age-related decline reflects changes in vagal tone, arterial stiffness, and cardiac remodeling that occur over time. It is not a failure. It is physiology.

Fitness level also shifts the baseline. Trained endurance athletes typically have higher resting HRV than sedentary individuals of the same age, reflecting greater vagal tone from chronic aerobic adaptation. A well-trained 40-year-old runner may have a higher RMSSD than an untrained 25-year-old. This makes cross-person comparisons even less meaningful.

The practical takeaway: population norms are less useful than your own trend. A 7-day rolling average smooths the daily noise — which is considerable — and reveals the directional shifts that actually matter. Your RMSSD dropping from a personal average of 42 to 31 over two weeks tells you something. Whether 42 is “good” compared to an Instagram influencer’s screenshot of 85 tells you nothing. The number itself matters less than its trajectory relative to your baseline.

When HRV-guided training works.

The research on HRV-guided training is genuinely promising — when the conditions are right. Plews et al. showed that elite endurance athletes who modulated training intensity based on daily HRV readings achieved better performance outcomes and fewer signs of overtraining compared to athletes following predetermined programs.^[2] Flatt demonstrated similar findings in team-sport athletes using smartphone-based HRV recording.^[4] Bellenger’s systematic review confirmed that autonomic heart rate regulation tracks meaningfully with training status across multiple athletic populations.^[5]

The pattern across these studies is consistent: HRV-guided training works when the athlete has a consistent measurement protocol, at least two weeks of baseline data to establish personal norms, and uses HRV as one input among several — alongside sleep quality, nutrition status, subjective wellness, and performance metrics. HRV is not a standalone oracle. It is one instrument in a panel. It gains predictive power in context and loses it in isolation.

SomaForge pulls HRV from HealthKit and plots it alongside sleep architecture and training load — because the number means nothing without the context around it.

When to ignore HRV.

HRV is a signal, not a command. It informs decisions. It does not make them. There are specific situations where the number is more noise than signal, and acting on it would be a mistake.

• •Single-day readings. One morning’s number is a data point, not a verdict. Day-to-day variability in RMSSD can be substantial even without any change in fitness or recovery status. Always interpret against your rolling average.
• •The day after alcohol. Alcohol suppresses HRV for 24–48 hours, even at moderate doses. This is well-documented and consistent. A tanked HRV the morning after drinks tells you nothing about your training readiness — it tells you that you drank.
• •During acute illness. Immune activation drives sympathetic dominance and suppresses HRV. If you have a cold or infection, your HRV will be low. This is your body doing its job, not a training signal.
• •When the number conflicts with how you feel AND you have an explanation. If your HRV says “recovered” but your legs are destroyed after yesterday’s volume PR, trust the legs. If your HRV says “suppressed” but you slept poorly because the dog barked at 3 AM and you otherwise feel fine, train.

The error most people make with HRV is treating it as an authority rather than an advisor. It is one signal among many. When it aligns with how you feel, your sleep data, and your recent training load, it adds confidence to the decision. When it conflicts with everything else, it is probably the outlier. Context determines the weight you give any single data point. That is true in medicine. It is true in training. It is true here.

References.